Clinical symptom nursing strategy intelligent recommendation system and method based on artificial intelligence

Abstract

The invention relates to a clinical symptom nursing strategy intelligent recommendation system and method based on artificial intelligence, and the method combines historical clinical symptoms and nursing big data to construct a corresponding nursing strategy intelligent recommendation model, thereby enabling the nursing strategy intelligent recommendation model to recognize the feature relationship between different symptom features and corresponding nursing strategies. And enabling the model to recognize the symptom characteristics of a clinical patient and intelligently recommend a corresponding nursing strategy. The system can also combine model output and nurse feedback, fuse and recommend a clinical symptom nursing strategy most related to patient symptoms, so that nurses can make preliminary assessment and deal with the appearing symptoms to confirm nursing problems, sort critical and urgent conditions, develop nursing measures, highlight nursing key points and improve the nursing efficiency. And potential complications can be prevented or treated in time. In clinical nursing work, clinical nurses can be effectively guided to perform judgment thinking and carry out predictive nursing, so that the nursing work is homogenized, and the nursing quality and safety are improved.

Description

Technical Field

[0001] The present disclosure relates to the field of intelligent nursing technology, and in particular to an artificial intelligence-based clinical symptom nursing strategy intelligent recommendation method, an artificial intelligence-based clinical symptom nursing strategy intelligent recommendation system and an electronic device. Background Art

[0003] Research shows that junior nurses have deficiencies in the content of condition observation, specialist nursing examinations, implementation and evaluation of nursing measures, and special medication observation and nursing. These problems may lead to untimely observation of patients' conditions and inadequate nursing, which may cause complications or even life-threatening conditions, requiring more medical resources for treatment, thereby increasing medical costs and posing the risk of potential medical disputes. Summary of the invention

[0004] In order to solve the above problems, the present application proposes an artificial intelligence-based clinical symptom nursing strategy intelligent recommendation method, an artificial intelligence-based clinical symptom nursing strategy intelligent recommendation system and an electronic device.

[0005] On the one hand, this application proposes an intelligent recommendation method for clinical symptom nursing strategies based on artificial intelligence, comprising the following steps: S1, read the patient electronic medical record data in the nursing electronic medical record system on the nurse station background; S2. extracting clinical symptom information of the patient from the electronic medical record data of the patient; S3, inputting the clinical symptom information into a preset nursing strategy intelligent recommendation model, identifying clinical nursing features in the clinical symptom information through the nursing strategy intelligent recommendation model and outputting a clinical symptom nursing strategy matching the clinical nursing features; S4, outputting the clinical symptom nursing strategy and writing it into the patient's electronic medical record data; S5. At the same time, the clinical symptom nursing strategy is recommended to the nursing end PDA / EDA.

[0006] As an optional implementation scheme of the present application, optionally, S2, extracting clinical symptom information of the patient from the electronic medical record data of the patient, comprises: Pre-build symptom keywords and symptom extraction logic; Combining the symptom keywords and the symptom extraction logic, constructing corresponding large language model prompt words; Inputting the large language model prompt word into a preset LLM large language model; Traversing the patient's electronic medical record data through the LLM large language model, identifying and extracting the corresponding clinical symptom information based on the large language model prompt words; Output the clinical symptom information to the nursing end PDA / EDA to remind the nursing staff to confirm and provide feedback.

[0007] As an optional implementation scheme of the present application, optionally, the method for generating the nursing strategy intelligent recommendation model includes: Collect clinical nursing big data, including clinical symptom data of several patients, information on the causes of symptoms, nursing conversation information on the causes of patients' illnesses by nursing staff, and clinical symptom nursing strategies based on the nursing conversation information; Performing feature engineering on the clinical nursing big data to extract clinical nursing features of the corresponding patients, including: symptom features and etiology feature data that trigger the corresponding symptom features, nursing dialogue text features, and nursing strategy text features; Preprocessing the clinical nursing features of each patient to construct a feature set consisting of a number of the clinical nursing features; Dividing the feature set into a training set and a validation set according to a preset ratio; Inputting the training set into a preset random forest model to perform feature training and learning, thereby constructing the intelligent recommendation model for the nursing strategy; The recognition performance of the nursing strategy intelligent recommendation model is verified using the verification set: If the verification is passed, the nursing strategy intelligent recommendation model is deployed to the nurse station background; If the verification fails, repeat the above steps to rebuild the nursing strategy intelligent recommendation model.

[0008] As an optional implementation scheme of the present application, optionally, S3, inputting the clinical symptom information into a preset nursing strategy intelligent recommendation model, identifying clinical nursing features in the clinical symptom information through the nursing strategy intelligent recommendation model and outputting a clinical symptom nursing strategy matching the clinical nursing features, including: Identify and output the symptom characteristics in the clinical symptom information, as well as the etiology characteristic data, nursing dialogue text characteristics and nursing strategy text characteristics associated with the symptom characteristics through the nursing strategy intelligent recommendation model; The etiology characteristic data is input into the nursing plan system on the backend, the nursing plan system recommends a nursing plan matching the etiology characteristic data to the nursing end PDA / EDA, and the nursing staff feeds back corresponding nursing plan selection information through the nursing end PDA / EDA; In combination with the nursing plan selection information and the nursing dialogue text features, the nursing question-and-answer system on the background recommends and outputs a number of corresponding nursing dialogues, and outputs each of the nursing dialogues to the nursing end PDA / EDA in order of priority, and the nursing staff feeds back corresponding nursing reply information through the nursing end PDA / EDA; In combination with the nursing response information and the nursing strategy text features, the strategy database on the backend recommends and outputs the corresponding clinical symptom nursing strategy.

[0009] On the other hand, the present application proposes an artificial intelligence-based clinical symptom nursing strategy intelligent recommendation system, which is used to implement the above-mentioned artificial intelligence-based clinical symptom nursing strategy intelligent recommendation method, including: Medical record reading module, used to read the patient electronic medical record data in the nursing electronic medical record system on the nurse station background; A symptom extraction module, used to extract clinical symptom information of the patient from the electronic medical record data of the patient; An intelligent recommendation module, used for inputting the clinical symptom information into a preset nursing strategy intelligent recommendation model, identifying clinical nursing features in the clinical symptom information through the nursing strategy intelligent recommendation model and outputting a clinical symptom nursing strategy matching the clinical nursing features; A strategy recording module, used for outputting the clinical symptom nursing strategy and writing it into the patient's electronic medical record data; The strategy push module is used to simultaneously recommend the clinical symptom nursing strategy to the nursing end PDA / EDA.

[0010] In another aspect, the present application further provides an electronic device, comprising: processor; a memory for storing processor-executable instructions; Wherein, the processor is configured to implement the artificial intelligence-based clinical symptom nursing strategy intelligent recommendation method when executing the executable instructions.

[0011] Technical effects of the present invention: The present invention is based on artificial intelligence and big data computing, combines historical clinical symptoms with nursing big data, and constructs a corresponding intelligent recommendation model for nursing strategies, so that the intelligent recommendation model for nursing strategies can identify the characteristic relationship between different symptom characteristics and corresponding nursing strategies. Let the model identify the symptom characteristics of clinical patients and intelligently recommend corresponding nursing strategies. The system can also combine the model output with the feedback of nursing staff, integrate and recommend clinical symptom nursing strategies that are most relevant to the patient's symptoms, and enable nurses to make preliminary assessments and deal with confirmed nursing problems for the symptoms that appear, sort the severity and urgency, launch nursing measures, highlight nursing priorities, and prevent or promptly deal with potential complications. In clinical nursing work, it can effectively guide clinical nurses to conduct critical thinking, carry out predictive nursing, homogenize nursing work, and improve nursing quality and safety.

[0012] Further features and aspects of the present disclosure will become apparent from the following detailed description of exemplary embodiments with reference to the attached drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] The accompanying drawings, which are incorporated in and constitute a part of the specification, illustrate exemplary embodiments, features, and aspects of the disclosure and, together with the description, serve to explain the principles of the disclosure.

[0014] Figure 1 It is shown as a schematic diagram of the implementation process of the present invention; Figure 2 Shown is an application system of the present invention; Figure 3 It is shown as a schematic diagram of the composition structure of the present invention; Figure 4 The application diagram of the electronic device of the present invention is shown as follows DETAILED DESCRIPTION

[0015] Various exemplary embodiments, features and aspects of the present disclosure will be described in detail below with reference to the accompanying drawings. The same reference numerals in the accompanying drawings represent elements with the same or similar functions. Although various aspects of the embodiments are shown in the accompanying drawings, the drawings are not necessarily drawn to scale unless otherwise specified.

[0016] The word “exemplary” is used exclusively herein to mean “serving as an example, example, or illustration.” Any embodiment described herein as “exemplary” is not necessarily to be construed as preferred or advantageous over other embodiments.

[0017] In addition, in order to better illustrate the present disclosure, numerous specific details are given in the following specific embodiments. It should be understood by those skilled in the art that the present disclosure can also be implemented without certain specific details. In some examples, means, components and circuits well known to those skilled in the art are not described in detail in order to highlight the main purpose of the present disclosure. Example

[0018] like Figure 1 As shown, on the one hand, the present application proposes an intelligent recommendation method for clinical symptom nursing strategies based on artificial intelligence, comprising the following steps: S1, read the patient electronic medical record data in the nursing electronic medical record system on the nurse station background; S2. extracting clinical symptom information of the patient from the electronic medical record data of the patient; S3, inputting the clinical symptom information into a preset nursing strategy intelligent recommendation model, identifying clinical nursing features in the clinical symptom information through the nursing strategy intelligent recommendation model and outputting a clinical symptom nursing strategy matching the clinical nursing features; S4, outputting the clinical symptom nursing strategy and writing it into the patient's electronic medical record data; S5. At the same time, the clinical symptom nursing strategy is recommended to the nursing end PDA / EDA.

[0019] The present invention trains and learns clinical nursing big data based on AI models. Through big data technology, feature engineering is performed on the clinical symptoms of several patients, the nursing dialogues caused by nurse intervention, and the clinical symptom nursing strategies adopted for the patient's history, the clinical nursing features of the corresponding patients are extracted and a feature set is constructed, and the AI ​​algorithm is used to train and construct a corresponding nursing strategy intelligent recommendation model. The model is subsequently deployed on the background server. When the patient's electronic medical record is constructed and the patient's symptoms are written, the system can read the patient's electronic medical record data, and input the clinical symptom information in the medical record data into the model. The model intelligently identifies the patient's clinical symptom characteristics, and outputs the corresponding clinical symptom nursing strategy and pushes it to the nursing staff, so that the nursing staff can refer to the nursing strategy recommended by the system and provide intelligent and automated nursing for the patient. Therefore, the patient's workload can be reduced and the nursing efficiency can be improved.

[0020] During the implementation process, various databases on the background can also be combined to conduct comprehensive strategy recommendation retrieval.

[0021] The big data involved in the model include clinical symptoms (such as dizziness, headache, chest tightness, shortness of breath, palpitations, chest pain, edema, fatigue, nausea and vomiting, abdominal pain, abdominal distension, diarrhea, etc.) and physical signs (abnormal heart rhythm, abnormal blood pressure, abnormal body temperature, abnormal breathing, abnormal urine volume, etc.), nursing problems (all related nursing problems matching the symptom), and each nursing problem includes related factors, nursing measures, observation and treatment of complications, health education, etc.

[0022] The backend server can manage the nursing terminals of each nursing staff, such as PDA or EDA nursing terminal devices. PDA or EDA can refer to the management plan of the existing nursing system.

[0023] The nursing plan system, nursing question and answer system and strategy database are deployed on the background server. The nursing plan system can store several nursing plans corresponding to different etiology characteristics; the nursing question and answer system can store a large number of nursing plans corresponding to different nursing plans and nursing dialogues containing different nursing dialogue text features (such as ICU nursing keywords); and the strategy database stores corresponding nursing response information and clinical symptom nursing strategies containing nursing strategy text features. The clinical symptom characteristics identified and output by the subsequent model can be input into the corresponding system or database, which will comprehensively retrieve and output the corresponding information based on the model output and the feedback information of the nursing staff, and finally recommend and output the corresponding clinical symptom nursing strategy.

[0024] Therefore, it is possible to integrate model predictions and nurse feedback to generate patient care strategies and improve the practicality and accuracy of clinical care.

[0025] like Figure 2 As shown, as an optional implementation scheme of the present application, optionally, S2, extracting the patient's clinical symptom information from the patient's electronic medical record data, includes: Pre-build symptom keywords and symptom extraction logic; Combining the symptom keywords and the symptom extraction logic, constructing corresponding large language model prompt words; Inputting the large language model prompt word into a preset LLM large language model; Traversing the patient's electronic medical record data through the LLM large language model, identifying and extracting the corresponding clinical symptom information based on the large language model prompt words; Output the clinical symptom information to the nursing end PDA / EDA to remind the nursing staff to confirm and provide feedback.

[0026] The present invention uses the LLM large language model on the background to assist sub-search. The LLM large language model can be an LLM large language model developed by the user based on GPT3 or the like, or a third-party LLM large language model (such as Baidu's Wenxinyiyan) requested through a background API call.

[0027] In order to integrate symptom keywords and symptom extraction logic into the application of the large language model (LLM), so as to effectively identify and extract clinical symptom information from the patient's electronic medical record data, the following steps can be referred to: 1. Establish symptom keywords and extraction logic First, a series of symptom keywords need to be identified, such as "headache", "fever", "cough", etc. These symptoms are information points that need special attention in electronic medical records. At the same time, a set of symptom extraction logic should be designed to help the model accurately identify the appearance of these symptoms in medical records, including direct symptom descriptions, doctors' diagnosis records, or patients' self-reports.

[0028] 2. Constructing large language model prompt words Based on the above symptom keywords and extraction logic, the prompt words of the large language model can be carefully designed. The prompt words should clearly guide the model on how to search and identify symptom-related information in the electronic medical record text.

[0029] Please pay special attention to the following symptom keywords: headache, fever, cough, etc. (You can add more keywords here as needed). Please ensure that your extraction results are comprehensive and accurate, covering all relevant clinical symptom information.

[0030] 3. Input the prompt word into the large language model Next, the carefully constructed prompt words are input into the preset LLM large language model. This usually involves using the prompt words as text data for the model input, and may require corresponding configuration according to the specific requirements of the model to ensure that the model can accurately understand and execute the instructions.

[0031] 4. Using large language models to process electronic medical record data With the help of the LLM large language model, the patient's electronic medical record data can be efficiently traversed. The model will intelligently search and identify information related to clinical symptoms in the medical record text according to the guidelines in the prompt words. This process may involve natural language processing (NLP) technologies such as text segmentation, part-of-speech tagging, named entity recognition, etc., to assist the model in more accurately locating symptom information.

[0032] 5. Extract and output results Finally, the large language model extracts all relevant clinical symptom information based on the prompt words and medical record data, and returns it in the expected format (such as text list, JSON object, etc.). This information can be directly used in subsequent medical analysis, diagnosis or treatment decisions.

[0033] like Figure 3 As shown, as an optional implementation scheme of the present application, optionally, the method for generating the nursing strategy intelligent recommendation model includes: Collect clinical nursing big data, including clinical symptom data of several patients, information on the causes of symptoms, nursing conversation information on the causes of patients' illnesses by nursing staff, and clinical symptom nursing strategies based on the nursing conversation information; Performing feature engineering on the clinical nursing big data to extract clinical nursing features of the corresponding patients, including: symptom features and etiology feature data that trigger the corresponding symptom features, nursing dialogue text features, and nursing strategy text features; Preprocessing the clinical nursing features of each patient to construct a feature set consisting of a number of the clinical nursing features; Dividing the feature set into a training set and a validation set according to a preset ratio; Inputting the training set into a preset random forest model to perform feature training and learning, thereby constructing the intelligent recommendation model for the nursing strategy; The recognition performance of the nursing strategy intelligent recommendation model is verified using the verification set: If the verification is passed, the nursing strategy intelligent recommendation model is deployed to the nurse station background; If the verification fails, repeat the above steps to rebuild the nursing strategy intelligent recommendation model.

[0034] In the traditional nursing process, when patients receive clinical care, nursing staff will collect and analyze the clinical symptom data of patients. The attending physician can determine and write the corresponding cause of the symptoms in the patient's medical record. Nursing staff can ask questions to patients based on clinical nursing requirements (such as scales), and can record and generate nursing dialogue information of the corresponding causes. At the same time, based on the nursing dialogue information between nurses and patients, nursing staff can generate corresponding clinical symptom nursing strategies (specific nursing measures, etc.) for patients.

[0035] Clinical nursing big data includes the above-mentioned information or data, which can be stored in a background database, and the administrator can retrieve and collect the clinical nursing data of the corresponding patients from the background database to construct clinical nursing big data. Subsequent feature engineering is performed to extract the corresponding clinical nursing features, and feature training is performed using an AI algorithm model to construct the nursing strategy intelligent recommendation model of the present invention.

[0036] The present invention can use machine learning algorithms to extract features, such as convolutional neural networks (CNNs). In order to use convolutional neural networks (CNNs) to extract features from clinical nursing big data, it is necessary to preprocess the data, construct a data format suitable for CNN input, and design a corresponding network structure to capture symptom features, etiology features, nursing conversation text features, and nursing strategy text features. The following is an example: 1. Data Preprocessing ‌Data cleaning‌: Remove irrelevant information, erroneous data and duplicate data from clinical nursing big data.

[0037] ‌Text segmentation‌: For nursing dialogue texts and nursing strategy texts, word segmentation processing is required to split the text into words or phrases.

[0038] ‌Feature Encoding‌: Symptom features and cause features: If these data are categorical, they can be converted into numerical features using One-Hot Encoding or Embedding technology.

[0039] Nursing conversation text features and nursing strategy text features: Word embedding technology (such as Word2Vec, GloVe, or BERT) is usually used to convert text into high-dimensional vector representation.

[0040] ‌Data Standardization / Normalization‌: For numerical features, standardization or normalization may be required to ensure that different features have similar scales.

[0041] 2. Build a CNN model CNN models are often used on image data, but can also be used on text data, especially when the text is converted into a matrix form similar to the image (for example, using word embeddings to convert the text into a two-dimensional matrix).

[0042] ‌Input layer‌: Receives preprocessed data. For text data, the input layer may be a two-dimensional matrix, where rows represent text (such as conversations or policies) and columns represent word embedding vectors.

[0043] ‌Convolutional layer‌: Use multiple convolution kernels to slide over the input data to extract local features. For text data, the convolution kernel can capture the features of n-gram (n consecutive words).

[0044] ‌Pooling layer‌: Downsample the output of the convolutional layer to reduce the data dimension while retaining important features. Commonly used pooling operations include Max Pooling and Average Pooling.

[0045] ‌Fully connected layer‌: The output of the pooling layer is flattened into a one-dimensional vector and passed through one or more fully connected layers for further feature extraction and classification.

[0046] ‌Output layer‌: Depending on the task requirements, the output layer can be a classifier (such as Softmax classifier) ​​for multi-classification tasks, or a regressor for prediction tasks.

[0047] 3. Model training and evaluation ‌Partitioning the dataset‌: Divide the clinical care big data into training set, validation set, and test set.

[0048] ‌Define loss function and optimizer‌: Select the appropriate loss function (such as cross entropy loss, mean square error, etc.) and optimizer (such as SGD, Adam, etc.) according to the task type (classification or regression).

[0049] ‌Model training‌: Use the training set data to train the CNN model and update the model parameters through the back-propagation algorithm.

[0050] ‌Model Evaluation‌: Evaluate the performance of the model on the validation set and test set, and use indicators such as accuracy, F1 score, and mean square error to measure the quality of the model.

[0051] ‌Model tuning‌: Adjust the model structure, hyperparameters, etc. based on the evaluation results to improve model performance.

[0052] 4. Feature extraction and application The trained CNN model can be used to extract features from clinical nursing big data. For new patient data, the same preprocessing steps as the training data are first performed, and then feature vectors are extracted through the forward propagation process of the CNN model. These feature vectors can be used for subsequent medical decision support, patient grouping, disease prediction and other tasks.

[0053] You can also use the LLM large language model to extract data features. For example: Using LLM (Large Language Model) to perform feature engineering on clinical nursing big data to extract clinical nursing features of patients is a complex but fruitful task. LLM large language models, such as GPT, BERT or their variants, have powerful natural language processing capabilities and can understand and analyze deep semantic information in text data. The following is a feature engineering framework based on LLM large language model, which aims to extract symptom features, etiology features, nursing dialogue text features and nursing strategy text features from clinical nursing big data: 1. Data preparation and preprocessing ‌Data Collection‌: Collect clinical nursing big data from hospital information systems, electronic medical record systems or nursing records.

[0054] ‌Data cleaning‌: Remove irrelevant information, erroneous data and duplicate records to ensure data accuracy and completeness.

[0055] ‌Text segmentation and standardization‌: Segment the nursing dialogue text and nursing strategy text, and unify the text format, such as converting to lowercase and removing punctuation marks.

[0056] 2. Use LLM large language model for feature extraction 2.1 Extraction of symptom characteristics and etiology characteristics ‌Construct prompt words‌: Design prompt words for symptoms and causes to guide the LLM model to focus on relevant information in the electronic medical record.

[0057] ‌Input Data‌: The preprocessed electronic medical record data together with the prompt words are input into the LLM model.

[0058] ‌Feature Extraction‌: Leverage the text understanding capabilities of the LLM model to extract symptom descriptions and corresponding etiology information from electronic medical records. This may involve natural language processing tasks such as named entity recognition (NER) or relation extraction.

[0059] 2.2 Nursing dialogue text feature extraction and nursing strategy text feature extraction ‌Conversation Understanding‌: Use the LLM model to understand the context of nursing conversations and identify key information, such as patient needs, nursing problems, and nursing measures.

[0060] ‌Strategy Analysis‌: Analyze the nursing strategy text and extract key elements, such as nursing goals, nursing measures, expected effects, etc.

[0061] ‌Text representation‌: Convert the nursing dialogue and nursing strategy text into a high-dimensional vector representation for subsequent feature analysis and model training. This can be achieved through the embedding layer of the LLM model or a specific text representation method (such as BERT's [CLS] token).

[0062] 3. Feature Integration and Representation ‌Feature fusion‌: The extracted symptom features, etiology features, nursing conversation text features, and nursing strategy text features are integrated to form a comprehensive clinical nursing feature set.

[0063] ‌Feature Representation‌: Depending on the task requirements, the features are represented in a format suitable for machine learning model input, such as vectors, matrices, or tensors.

[0064] 4. Model training and application ‌Model selection‌: Select a suitable machine learning model based on the results of feature engineering and the task type (such as classification, regression, clustering, etc.).

[0065] ‌Model training‌: Use the extracted features to train the machine learning model and optimize the model parameters to improve the prediction performance.

[0066] The specific method can be selected by the administrator.

[0067] During model training, the present invention uses a random forest model to perform feature training and learning on the clinical nursing features in the training set to construct the nursing strategy intelligent recommendation model. Specifically, the principle and application of the random forest model can be combined, for example: 1. Data Preparation ‌Training set data‌: Ensure that the training set data (for example, divide the feature set into training set and validation set in an 8:2 ratio) has been cleaned and preprocessed, and contains all necessary features, such as symptom features, etiology features, nursing conversation text features (which may have been converted into numerical features or vector representations), and corresponding nursing strategy labels.

[0068] 2. Random Forest Model Initialization ‌Select parameters‌: According to the characteristics of the dataset and task requirements, select the parameters of the random forest model, such as the number of trees, maximum depth, minimum number of sample splits, etc.

[0069] ‌Model Initialization‌: Initialize the random forest model with the selected parameters.

[0070] 3. Feature training and learning ‌Input training set‌: Input the prepared training set data into the random forest model.

[0071] ‌Model training‌: The random forest model will be trained based on the input features and labels, and the accuracy and stability of the model will be improved by building multiple decision trees and integrating their prediction results.

[0072] ‌Feature Importance Evaluation‌: During or after training, you can evaluate the importance of individual features to understand which features contribute most to the model predictions.

[0073] 4. Model evaluation and optimization ‌Cross-Validation‌: Use cross-validation techniques (such as K-fold cross-validation) to evaluate the performance of the model and ensure that the model performs well on unseen data.

[0074] Hyperparameter tuning: Based on the results of cross-validation, the hyperparameters of the random forest model are adjusted to further optimize the model performance.

[0075] ‌Performance evaluation‌: Use appropriate evaluation metrics (such as accuracy, F1 score, AUC, etc.) to quantify the performance of the model. This is done by the administrator. The model file can then be loaded and deployed on the backend.

[0076] 5. Model construction and storage ‌Build the final model‌: After finding the optimal parameter combination, retrain the random forest model using all training set data to build the final nursing strategy intelligent recommendation model.

[0077] ‌Model saving‌: Save the trained model to a file or database for subsequent loading and prediction.

[0078] 6. Model deployment and application ‌Model deployment‌: Deploy the trained random forest model to the corresponding application environment, such as hospital information system or nursing decision support system.

[0079] ‌Intelligent Recommendation‌: When new patient data is input, the model is able to predict and recommend corresponding care strategies based on the extracted features.

[0080] ‌Continuous Monitoring and Updates‌: Continuously monitor the performance of the model in real-world applications and update the model as needed to adapt to new data or changes in nursing practice.

[0081] Through the above steps, the random forest model can be used to build an intelligent nursing strategy recommendation model, which can predict and recommend the most appropriate nursing strategy based on the patient's clinical nursing characteristics.

[0082] As an optional implementation scheme of the present application, optionally, S3, inputting the clinical symptom information into a preset nursing strategy intelligent recommendation model, identifying clinical nursing features in the clinical symptom information through the nursing strategy intelligent recommendation model and outputting a clinical symptom nursing strategy matching the clinical nursing features, including: Identify and output the symptom characteristics in the clinical symptom information, as well as the etiology characteristic data, nursing dialogue text characteristics and nursing strategy text characteristics associated with the symptom characteristics through the nursing strategy intelligent recommendation model; The etiology characteristic data is input into the nursing plan system on the backend, the nursing plan system recommends a nursing plan matching the etiology characteristic data to the nursing end PDA / EDA, and the nursing staff feeds back corresponding nursing plan selection information through the nursing end PDA / EDA; In combination with the nursing plan selection information and the nursing dialogue text features, the nursing question-and-answer system on the background recommends and outputs a number of corresponding nursing dialogues, and outputs each of the nursing dialogues to the nursing end PDA / EDA in order of priority, and the nursing staff feeds back corresponding nursing reply information through the nursing end PDA / EDA; In combination with the nursing response information and the nursing strategy text features, the strategy database on the backend recommends and outputs the corresponding clinical symptom nursing strategy.

[0083] When applying the model, the nursing strategy intelligent recommendation model can be used to identify and output the symptom features in the clinical symptom information, as well as the etiology feature data, nursing dialogue text features and nursing strategy text features associated with the symptom features. After identifying and outputting each feature, the system can also recommend the corresponding features of the patient to the corresponding system or database, so as to retrieve and output the corresponding etiology, nursing dialogue and clinical symptom nursing strategy to the PDA for the nursing staff to refer to and select to output the corresponding information, and then combine the nursing staff's selection of the current feature and the next feature to comprehensively evaluate the next nursing information. Therefore, the model can assist in predicting the cause of the patient's symptoms, and the system combines the nursing staff's judgment of the cause, comprehensively outputs the corresponding cause evaluation results, and further evaluates the nursing dialogue. Therefore, it is possible to comprehensively combine the model prediction results and the nursing staff's selection of the results, and continue to intervene and adjust the prediction results of the next stage, so as to improve the accuracy of strategy recommendation.

[0084] For example, take chest tightness as an example to explain in detail. The nursing electronic medical record system has a built-in symptom database (here, taking chest tightness as an example, it is a chest tightness database that stores different types of chest tightness symptom data), which contains all possible causes of chest tightness symptoms and the corresponding nursing dialogues for each cause and the corresponding nursing strategies for each nursing dialogue: 1) Symptom Database 1. Respiratory system: Ventilation disorders: a. Obstructive ventilation disorders: airway stenosis, spasm, sputum, blood clots, foreign bodies, status asthmaticus, laryngeal edema, tumor compression, and OSAHS; b. Restrictive ventilation disorders: suppression of the respiratory center, myasthenia gravis, pleural effusion, pneumothorax, abdominal distension, diaphragm elevation, chest deformity, pleurisy, pain, tension and anxiety; (2) Ventilation disorders: pulmonary edema, pulmonary consolidation, atelectasis, pulmonary embolism, ARDS, COPD, DIC, and intrapulmonary arteriovenous fistula; (3) Gas utilization disorder: decreased hemoglobin and carbon monoxide poisoning.

[0085] 2. Circulatory system: arrhythmia, angina pectoris, myocardial infarction, heart failure, acute pulmonary edema, atrial septal defect, ventricular septal defect, rheumatic heart disease, etc.

[0086] 3. Digestive system: intestinal obstruction, constipation, intestinal flatulence, esophageal stenosis, reflux esophagitis, ascites due to cirrhosis, cholecystitis, pancreatitis, etc.

[0087] 4. Mental aspect: caused by mental and psychological factors, such as depression, bad mood, etc.

[0088] II) Causes When a clinical patient complains of chest tightness, the nurse conducts a preliminary assessment and enters "chest tightness" in the symptom box in the nursing plan system of the nursing electronic medical record for retrieval. The first box that appears is the content of the preliminary assessment and treatment: Vital signs (temperature, pulse, blood pressure, respiration, oxygen saturation, pain); Auscultate breath sounds; Duration of chest tightness; The patient's complexion and mental status; The state before the onset of symptoms (such as eating, coughing, infusion, etc.); Report to doctor.

[0089] After the nurse enters the patient's information item by item, if the vital signs are abnormal and meet the hospital's indications for calling a critical care doctor, a dialog box will pop up with suggestions and instructions to continue to pay attention to the patient's condition. If not, the interface will pop up a box for further evaluation and treatment: (Whether double-clicking can directly retrieve the medical record) diagnosis; Medical history; Past history; Surgical history; CXR or CT Blood results: such as DDI, blood gas, white blood cells, BNP and other abnormal results; B-ultrasound: cardiac ultrasound and other abnormal ultrasounds; Electrocardiogram (ECG) Current medications; The amount of people entering and leaving the class; Whether the drainage tube is unobstructed and the drainage volume; Other unusual symptoms; Other abnormal physical signs; What risks exist (such as pressure ulcers, falls, VTE, double-click to automatically retrieve).

[0090] 3) Nursing dialogue: By inputting the above information, the system automatically identifies and screens, and locks in the most likely causes of the patient's chest tightness. At the same time, if the evidence is insufficient, the system pops up a dialog box and recommends further examination. If pulmonary embolism is considered, further examination of CTPA (pulmonary angiography) is required. After confirming the main cause of the patient's chest tightness, the system automatically pops up a dialog box for nursing questions based on all the information currently entered, and sorts them from major to minor issues. For example: Impaired gas exchange; Ineffective airway clearing; Heart rhythm disorders There is a risk of infection; Electrolyte imbalance; Impaired skin integrity.

[0091] IV) Clinical symptom nursing strategy: After clicking each nursing question, all the relevant nursing information of the nursing question will appear. The nurse will check the expected goals, relevant factors, relevant nursing measures, key points of observation, potential complications and health education according to the specific situation of the patient, which is in line with the patient's treatment and care. At the same time, click on the complication, and the drop-down menu will display the measures to prevent the complication and the response to the complication. The following takes the nursing question of impaired gas exchange as an example.

[0092] 1. Expected goals Refers to oxygen saturation > 95%; The patient was alert and had no abnormal changes in mental status; Respiratory rate 16-20 times / min; Hemoglobin was within the normal range; No dyspnea, tachypnea, or cyanosis; BP and P were within normal range; Blood gas analysis was within normal range; Regular breathing rhythm; The breath sounds of both lungs are clear and symmetrical; 2. Related factors Airway restriction; Decreased alveolar surfactant; Lung Qi and blood ratio imbalance; Limited mechanical expansion of the chest wall; Weakness of respiratory muscles 3. Observation Activity tolerance; Other causes of impaired gas exchange, such as the degree of pain; Breath sounds in both lungs; Blood oxygen saturation; 4. Blood gas analysis Monitor consciousness, blood pressure, circulation, and note whether restlessness or irritability worsens; Lip and nail bed color; Respiratory rhythm, frequency, depth; Swelling and bleeding of tissues around the tracheotomy; 5. Treatment Keep the airway open, exhale sputum in time, and increase oxygen intake before and after clearing the airway; Use antibiotics correctly as prescribed by your doctor; Assist doctors with atrial fibrillation; Give the patient a position that is conducive to breathing, such as semi-recumbent position or high-pillow position; For patients with rib fractures, assist the doctor in chest fixation or make preparations for surgery if necessary; Give oxygen or mechanically assisted pressurized ventilation according to doctor's orders; Give analgesics according to doctor's orders, and use sedatives and respiratory depressants with caution; Nebulizer inhalation dilutes sputum; Turning over every 2 hours, chest physical therapy; 6. Mission The importance of early postoperative activities, and teaching appropriate sleeping methods and deep breathing methods; After taking analgesics, encourage patients to cough and expectorate; Improve nutrition by eating high-protein, high-calorie, high-vitamin foods; Ensure adequate rest and provide a comfortable position; Potential complications; respiratory failure; asphyxia.

[0093] Therefore, this application is based on artificial intelligence and big data computing, combined with historical clinical symptoms and nursing big data, to build a corresponding intelligent recommendation model for nursing strategies, so that the intelligent recommendation model for nursing strategies can identify the characteristic relationship between different symptom characteristics and corresponding nursing strategies. Let the model identify the symptom characteristics of clinical patients and intelligently recommend corresponding nursing strategies. This system can also combine the model output with the feedback of nursing staff to integrate and recommend clinical symptom nursing strategies that are most relevant to the patient's symptoms. Nurses can make preliminary assessments and deal with nursing problems based on the symptoms that appear, sort the severity and urgency, launch nursing measures, highlight nursing priorities, and prevent or promptly deal with potential complications. In clinical nursing work, it can effectively guide clinical nurses to conduct critical thinking, carry out predictive nursing, homogenize nursing work, and improve nursing quality and safety.

[0094] Obviously, those skilled in the art should understand that the implementation of all or part of the processes in the above-mentioned embodiments can be completed by instructing the relevant hardware through a computer program, and the program can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the above-mentioned embodiments. Those skilled in the art can understand that the implementation of all or part of the processes in the above-mentioned embodiments can be completed by instructing the relevant hardware through a computer program, and the program can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the above-mentioned embodiments. Among them, the storage medium can be a disk, an optical disk, a read-only memory (ROM), a random access memory (RAM), a flash memory (Flash Memory), a hard disk (Hard Disk Drive, abbreviated as: HDD) or a solid-state drive (SSD), etc.; the storage medium can also include a combination of the above-mentioned types of memory. Example

[0095] Based on the implementation principle of Example 1, on the other hand, the present application proposes an artificial intelligence-based clinical symptom nursing strategy intelligent recommendation system, which is used to implement the above-mentioned artificial intelligence-based clinical symptom nursing strategy intelligent recommendation method, including: Medical record reading module, used to read the patient electronic medical record data in the nursing electronic medical record system on the nurse station background; A symptom extraction module, used to extract clinical symptom information of the patient from the electronic medical record data of the patient; An intelligent recommendation module, used for inputting the clinical symptom information into a preset nursing strategy intelligent recommendation model, identifying clinical nursing features in the clinical symptom information through the nursing strategy intelligent recommendation model and outputting a clinical symptom nursing strategy matching the clinical nursing features; A strategy recording module, used for outputting the clinical symptom nursing strategy and writing it into the patient's electronic medical record data; The strategy push module is used to simultaneously recommend the clinical symptom nursing strategy to the nursing end PDA / EDA.

[0096] Please understand the functions and interactions of the above modules in conjunction with the corresponding steps in Example 1.

[0097] The modules or steps of the present invention described above can be implemented by a general-purpose computing system, they can be concentrated on a single computing system, or distributed on a network composed of multiple computing systems, and optionally, they can be implemented by a program code executable by a computing system, so that they can be stored in a storage system and executed by the computing system, or they can be made into individual integrated circuit modules, or multiple modules or steps therein can be made into a single integrated circuit module for implementation. Thus, the present invention is not limited to any specific combination of hardware and software. Example

[0098] like Figure 4 As shown, further, in another aspect, the present application also proposes an electronic device, including: processor; a memory for storing processor-executable instructions; Among them, the processor is configured to implement the intelligent recommendation method for clinical symptom nursing strategies based on artificial intelligence described in Example 2 when executing the executable instructions.

[0099] The electronic device of the embodiment of the present disclosure includes a processor and a memory for storing instructions executable by the processor. The processor is configured to implement the clinical symptom nursing strategy intelligent recommendation method based on artificial intelligence described in the above embodiment 2 when executing the executable instructions.

[0100] Here, it should be noted that the number of processors can be one or more. At the same time, the electronic device of the embodiment of the present disclosure may also include an input system and an output system. Among them, the processor, memory, input system and output system may be connected through a bus or in other ways, which are not specifically limited here.

[0101] The memory, as a computer-readable storage medium, can be used to store software programs, computer executable programs and various modules, such as the program or module corresponding to the clinical symptom nursing strategy intelligent recommendation method based on artificial intelligence in the embodiment of the present disclosure. The processor executes various functional applications and data processing of the electronic device by running the software programs or modules stored in the memory.

[0102] The input system can be used to receive input numbers or signals. The signal can be a key signal related to user settings and function control of the device / terminal / server. The output system can include display devices such as display screens.

[0103] The embodiments of the present disclosure have been described above, and the above description is exemplary, not exhaustive, and is not limited to the disclosed embodiments. Many modifications and changes will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The selection of terms used herein is intended to best explain the principles of the embodiments, practical applications, or technical improvements in the market, or to enable other persons of ordinary skill in the art to understand the embodiments disclosed herein.

Claims

1. An artificial intelligence-based intelligent recommendation method for clinical symptom nursing strategies, characterized in that: The steps include: S1, read the patient electronic medical record data in the nursing electronic medical record system on the nurse station background; S2. extracting clinical symptom information of the patient from the electronic medical record data of the patient; S3, inputting the clinical symptom information into a preset nursing strategy intelligent recommendation model, identifying clinical nursing features in the clinical symptom information through the nursing strategy intelligent recommendation model and outputting a clinical symptom nursing strategy matching the clinical nursing features; S4, outputting the clinical symptom nursing strategy and writing it into the patient's electronic medical record data; S5. At the same time, the clinical symptom nursing strategy is recommended to the nursing end PDA / EDA.

2. The method for intelligent recommendation of clinical symptom nursing strategies based on artificial intelligence according to claim 1 is characterized in that: S2. Extracting clinical symptom information of the patient from the electronic medical record data of the patient, including: Pre-build symptom keywords and symptom extraction logic; Combining the symptom keywords and the symptom extraction logic, constructing corresponding large language model prompt words; Inputting the large language model prompt word into a preset LLM large language model; Traversing the patient's electronic medical record data through the LLM large language model, identifying and extracting the corresponding clinical symptom information based on the large language model prompt words; Output the clinical symptom information to the nursing end PDA / EDA to remind the nursing staff to confirm and provide feedback.

3. The method for intelligently recommending clinical symptom nursing strategies based on artificial intelligence according to claim 1 is characterized in that: The method for generating the nursing strategy intelligent recommendation model comprises: Collect clinical nursing big data, including clinical symptom data of several patients, information on the causes of symptoms, nursing conversation information on the causes of patients' illnesses by nursing staff, and clinical symptom nursing strategies based on the nursing conversation information; Performing feature engineering on the clinical nursing big data to extract clinical nursing features of the corresponding patients, including: symptom features and etiology feature data that trigger the corresponding symptom features, nursing dialogue text features, and nursing strategy text features; Preprocessing the clinical nursing features of each patient to construct a feature set consisting of a number of the clinical nursing features; Dividing the feature set into a training set and a validation set according to a preset ratio; Inputting the training set into a preset random forest model to perform feature training and learning, thereby constructing the intelligent recommendation model for the nursing strategy; The recognition performance of the nursing strategy intelligent recommendation model is verified using the verification set: If the verification is passed, the nursing strategy intelligent recommendation model is deployed to the nurse station background; If the verification fails, repeat the above steps to rebuild the nursing strategy intelligent recommendation model.

4. The method for intelligently recommending clinical symptom nursing strategies based on artificial intelligence according to claim 1 is characterized in that: S3, inputting the clinical symptom information into a preset nursing strategy intelligent recommendation model, identifying clinical nursing features in the clinical symptom information through the nursing strategy intelligent recommendation model and outputting a clinical symptom nursing strategy matching the clinical nursing features, including: Identify and output the symptom characteristics in the clinical symptom information, as well as the etiology characteristic data, nursing dialogue text characteristics and nursing strategy text characteristics associated with the symptom characteristics through the nursing strategy intelligent recommendation model; The etiology characteristic data is input into the nursing plan system on the backend, the nursing plan system recommends a nursing plan matching the etiology characteristic data to the nursing end PDA / EDA, and the nursing staff feeds back corresponding nursing plan selection information through the nursing end PDA / EDA; In combination with the nursing plan selection information and the nursing dialogue text features, the nursing question-and-answer system on the background recommends and outputs a number of corresponding nursing dialogues, and outputs each of the nursing dialogues to the nursing end PDA / EDA in order of priority, and the nursing staff feeds back corresponding nursing reply information through the nursing end PDA / EDA; In combination with the nursing response information and the nursing strategy text features, the strategy database on the backend recommends and outputs the corresponding clinical symptom nursing strategy.

5. An artificial intelligence-based clinical symptom nursing strategy intelligent recommendation system, used to implement the artificial intelligence-based clinical symptom nursing strategy intelligent recommendation method according to any one of claims 1 to 4, characterized in that: include: Medical record reading module, used to read the patient electronic medical record data in the nursing electronic medical record system on the nurse station background; A symptom extraction module, used to extract clinical symptom information of the patient from the electronic medical record data of the patient; An intelligent recommendation module, used for inputting the clinical symptom information into a preset nursing strategy intelligent recommendation model, identifying clinical nursing features in the clinical symptom information through the nursing strategy intelligent recommendation model and outputting a clinical symptom nursing strategy matching the clinical nursing features; A strategy recording module, used for outputting the clinical symptom nursing strategy and writing it into the patient's electronic medical record data; The strategy push module is used to simultaneously recommend the clinical symptom nursing strategy to the nursing end PDA / EDA.

6. An electronic device, characterized in that include: processor; a memory for storing processor-executable instructions; Wherein, the processor is configured to implement the artificial intelligence-based clinical symptom nursing strategy intelligent recommendation method as described in any one of claims 1-4 when executing the executable instructions.

Images